the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 15 Sep 2026
Mainz Convective Transport and Scavenging (MCTS version 1.0) – Introduction and Evaluation of the new Column Model for Convection-Chemistry-Interactions
Abstract. Convective systems can rapidly redistribute atmospheric trace gases. For soluble species, not only does convective transport play a key role, but also the interactions with cloud hydrometeors, i.e. scavenging. Global chemistry-climate models have difficulties reproducing observed vertical profiles of soluble trace gases under convective influence. One example is the model EMAC (ECHAM MESSy Atmospheric Chemistry). There are two possible reasons for this issue: (1) the lack of retention and multi-phase transport in the model and (2) the subsequent treatment of the convective transport and the convective scavenging. We present the new column model Mainz Convective Transport and Scavenging (MCTS), which overcomes these shortcomings. MCTS considers transport of trace substances and their scavenging via the ice and liquid hydrometeors quasi-simultaneously. Furthermore, the new model takes retention and multi-phase transport into account. MCTS is evaluated against aircraft observations from the NASA SEAC4RS campaign. WRF-Chem cloud-resolving output was used as meteorological input data for MCTS to accurately represent the convective storm. The new column model performs reasonably well given that it was originally designed for the application in a global model and not for the inter-comparison with a single convective storm. Additionally, MCTS is compared to the convective transport and scavenging in the chemistry-climate model EMAC. Both models affect insoluble species similarly, but have a different effect on the soluble species. Sensitivity simulations with MCTS demonstrate that the time scales of convective transport and the dissociation reactions within the droplet have a major effect on the simulated tracer profiles. MCTS can represent the competition between these two process types, which is a benefit for the calculation of vertical profiles of soluble tracer profiles compared to EMAC. MCTS opens the path for a wide range of future studies, for example, sensitivity studies regarding the ice phase scavenging, which is fraught with high uncertainty. In the long run, we aim to improve the representation of convective transport and convective scavenging in global chemistry-climate models with MCTS.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Geoscientific Model Development.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.- Preprint
(2062 KB) - Metadata XML
-
Supplement
(205 KB) - BibTeX
- EndNote
Status: open (until 01 Dec 2026)
- AC1: 'Comment on egusphere-2026-4821', Adrienne Jeske, 23 Sep 2026 reply
-
RC1: 'Comment on egusphere-2026-4821', Anonymous Referee #1, 23 Oct 2026
reply
Review of ‚Mainz Convective Transport and Scavenging (MCTS version 1.0) – Introduction and Evaluation of the new Column Model for Convection-Chemistry-Interactions’ by Adrienne Jeske et al., submitted to GMD.
The paper describes and evaluates MCTS, a column model designed to treat chemical interactions, mixing and multiphase transport of trace gases and aerosol constituents. Apart from transport processes in different classes (air, liquid and frozen hydrometeors, entrainment/ detrainment) it particular covers exchange processes between gas phase and cloud droplets, rain droplets, ice and snow crystals. Therefore, scavenging and retention, and very simplified also aqueous phase chemistry is explictly treated. The model is written in a way that each process (incl. transport) is solved by ordinary differential equations similar to chemical mechanism solvers. The KPP solver is used.
The model is tested in two setups. The first is using storm-resolving simulations with WRF-Chem as input and comparing it to aircraft observations in the inflow and outflow of the very same storm that was simulated with WRF-Chem (provided by a previous study). The second setup compares MCTS to the global chemistry transport model EMAC and its submodels for convective transport and scavenging.
Overall, the model seems to compare well against the observations, i.e. mostly evaluated as scavenging efficiencies between inflow and outflow of the convective storm. It could by improved by slowing down dissociation reactions for e.g. SO2. However, it was not easy to understand which data was used for input and initialization of the model and which is used for validation. The description in the manuscript should be revised to make this clearer. Further, model results in the storm-resolving setup showed some persistent spikes, which were not discussed and which I could not understand right away.
Compared to EMAC, MCTS performs similar in my view. It is discussed that this is expected for insoluble species, however, for soluble species in my view the changes introduced by MCTS are not substantial. The discussion on this matter is in my view sometimes a bit misleading and should be revised.
At the end of the paper, a note on the computation time is given. It would be great if a quantitative implication for a global model run or a typical global model time step could be provided.
Overall, the model description and provided supplementary material is generally well written and provides most necessary information. Nevertheless, there were a few occasions where the description confused me.
The manuscript can be accepted after minor revision.
Major comments:
l. 15-16: This sounded interesting to me when I first read the abstract. Later comparison of current EMAC vs. MCTS however gives the impression that the two realizations are rather similar to each other and MCTS would not have a substantial benefit.
l. 640-644: The good agreement between EMAC and MCTS reads somewhat that implementing MCTS does not provide much benefit for EMAC, despite the different processes covered. I think here or in the conclusion, a paragraph on the proven benefits of using MCTS instead of the current EMAC submodels in this should be better highlighted and discussed.
l. 674: I wouldn't agree that there significant differences between MCTS and EMAC realization also for the soluble species. The profiles look very similar (which you also write in the next sentence) and the reported SEs deviate only by a few percent. Also the wording signifcant deviates from the thorough description given in sec 5.3. Which aspects does 'significant' refer to in this sentence? Maybe, this statement should be written less general.
l. 676: How do you evaluate MCTS is superior to EMAC? As far as I understand, you could not compare EMAC and MCTS against observations? A few sentences later you also state that it is not possible to conclude which model performs better.
l. 180, l. 339, l. 700-703: In l. 180 you state that input data was taken from EMAC and WRF-Chem output. But later (l. 339) it is written that DC8 observations were used for the background, which is used as initialization? In l. 342, it is stated ‚S(DC8) is used to validate the performance of MCTS.‘ Can you please disentangle better which dataset is used as input data, i.e. what are the background profiles (DC8 obs or WRF output) seen in the Fig. 5 and further, and what data is used for evaluation/ validation?
I guess the 'background' is cloud-free and used for input of MCTS tracers and WRF-Chem Output data used for input of the vertical transport?
So, the inflow and outflow observations are used for observations, right? Or did you also intend to evaluate the convective simulations against the cloud-free background profiles (which presumably are used as input)? I would see the latter quite critical.
The current data and setup description in section 3 and 4 leaves me quite confused on this aspect and it should be revised to more concretely state which datasets are used for input and which for validation. A revised clarification would improve the manuscript.
Figure 5 - 8, l. 426-425: Why are there these spikes / peaks at certain vertical levels? You state that they are ‚highly correlated with the inflow and outflow pattern of the multicell storm, i.e. with organised entrainment and detrainment.’ However, I don’t understand why these should necessarily leads to these peaks?
The peaks seem quite uniformly distributed. To a modeller's eye this directly looks suspicious and somehow like an artefact/ accumulation of fluxes at certain levels. In particular as they stay at the same locations at 10 and 30min. Can you demonstrate more what causes these? Are they depending on the choice of vertical grid resolution / number of levels?
Minor comments:
l. 4, ‚two possible reasons‘: Isn’t multiphase chemistry also a possible reason?
l. 14-15: The sensitivity on the dissociation time scale is done for the convection resolving WRF-Chem data, but is presented in the abstract directly referring to EMAC. This is somewhat misleading in my view.
l. 34-35: Since it is the introduction, it would be great for readers if you could shortly mention the reason(s) of the findings of Lawrence et al., (2003)?
l. 60-61: Would the findings of Mari et al., (2000) on enhanced H2O2 in the outflow hold if reactive uptake would be considered? As H2O2 rapidly oxidizes SO2 to sulfate in the aqueous phase, its reactive uptake leads to a strong depletion of H2O2 in the gas phase after a cloud passage. At least in the troposphere.
l. 61: Why is efficient scavenging of HNO3 indicating the impact of retention?
l. 72-73: Shortly providing the key result of the obs - model comparison of the retention factor by Cuchiara et al. (2023) would be nice instead of just stating that they did this comparison.
l. 86: Can you described here or later in chapter 2 more about how CVTRANS (which MCTS is based on) calculates the convective transport? Does it calculate vertical advection e.g., using subgrid-scale assumptions for vertical wind or similar?
l. 89: What is the bulk approach with reference to convective transport? I think, it would be helpful to shortly explain it.
l. 94: What is CONVECT? Also here, a short hint would help a less interrupted (due to checking the references) read.
l. 101-109: I think, aqueous phase reactions towards sulfate should be outlined as a source of uncertainty as well. Further, the introduction discusses the potential effects of retention quite frequent, however, no investigations on retention follow in manuscript but it is speculatively referred to its effects later in the discussion of the results.
l. 129: It is not evaluated if MCTS performs better than the EMAC submodels. This is also stated later. Therefore, mentioning it as a motivation for the comparison is misleading.
l. 168: What does 'calculates the convective transport' mean? Do you mean vertical advection? Is this provided by CONVECT? How is the subgrid-scale treated? While it is totally fine referring to previous publications, it would be nice to provide a few key information in the manuscript.
Fig. 1 and l. 241: Do n, n+1, etc. define the grid cell center points or edges? Typically, in 3D CTMs Mass fluxes are usually defined at the grid cell edges, whereas concentrations and other scalars are defined at the grid cell centers.
l. 256: The locations updraft, downdraft, and environment are here defined as ‚reservoirs‘. However, later the term ‚reservoir‘ is also used for masses in, e.g., different condensate categories (e.g., l. 293). I found this a bit distracting and lead to misunderstandings. You could either make the definition of ‚reservoirs‘ less strict/ narrow or use a different term for other usages.
l. 263: That the model does not consider gas phase chemistry should be mentioned earlier when referring to chemical mechanisms. This was not clear to me until this point.
l. 287-288: It would help understanding the content of the model better, if you could provide in 1-2 sentences, what the ‚Scm‘ mechanism covers?
l. 305-306: Can give more detail about what 'did not perform well' means? Over- / underestimated, compared against observations, …? If so, which ones?
l. 315: 'some kind of' is rather colloquial. Maybe more concretely express how the surface layers of ice are treated / what is assumed?
l. 329-331: Does this mean that MCTS calculates mass transfer rates between hydrometeor categories? Then it should be described in a bit more detail how this is done.
l. 342: It is not said what core A refers to. I was a bit distracted from the naming core 7 and core A. Maybe, you can make it clearer in the text.
l. 348: Is ‚model top‘ referring to MCTS?
l. 350: It is not mentioned which vertical grid is actually chosen for S(DC8).
l. 355: Can you comment on the lower borders of the updraft and downdraft velocities for sampling? To which extent can it be expected that the model result is sensitive to the choice of these parameters?
l. 456: Does the mentioning of ‚high vertical variability’ refer to the spikes? It should be further explained and this statement contextualized according to the response to the comment on these spikes given in the major comments.
Figure 6 and according text: I could not understand precisely what is the difference between MCTS 30 min and MCTS 30 min all? Is the latter only updraft region? Then it should be mentioned clearer that MCTS 30 min is not only in the updraft region.
l. 477: Reference to Bela et al. (2016) and other papers on Scavenging Efficiency: Mention if these are field measurements or model experiments.
l. 498-499: Does retention in MCTS still play a role if the ice water content is zero, i.e. at the model top level. I assumed that in this case all tracers trapped in or attached to ice are released back to the gas phase?
l. 500-501: In the sentence before you state that retention, at least on/ off leads to a considerable effect on H2O2. Maybe rephrase to make it easier to understand.
Table 4 & 5: I like this overview. However, there are many further studies mentioned in the text, but not given in the tables. I would suggest to include these in the table. If so, they could be sorted between model estimates and field observations / measurements.
l. 541-543: In that case, I assume their responses should be similar in relative terms?
l. 552-554, table S3: Maybe mention in the text if dissociations are slowed down or enhanced. From the table S3 I assume that most reactions were slowed down. But I'm puzzled by the two numbers that are given for many species. I'd read 1.E-6 x 1.E3 as equal to 1.E-3? If so, why not giving this number for easier readability?
l. 579-580: Isn't it rather the low vertical wind velocities / mass flux compared to the WRF-Chem simulation? I.e. they can maybe not be expected to alter the profiles much in absolute concentration terms.
l. 638: Can you explain the presumed effect of retention and the conclusion a bit more, please? As stated later, I understand a sensitivity study (which would be great) is kept for future research.
l. 650: This is not very concrete. What inaccuracies did you want to address?
l. 651-652: Could you write more concretley, which uncertainty you are referring to?
l. 668-669: The justification in brackets should be mentioned already when the sensitivity experiment is introduced in sec. 5.2.
l. 684-685: Can you scale up this computation time estimate to a global model simulation? E.g., if it is executed in each grid cell always needs 10s then each times tep would take X seconds longer?
l. 707-708: I suggest to state that this refers to slower dissociation of SO2.
Further remarks, typos, spelling,…
l. 17: As a non-native speaker, I am not familiar with the word ‚fraught’. Maybe in the abstract a simpler wording could be used.
l. 24-25: Isn't 'moistening of the lower stratosphere' the same as 'lower stratospheric water vapour concentration'?
l. 61: Retention is used the first time here. The explanation given in l. 71 should be given here already.
l. 62: inter alia is unusual -> amongst others?
l. 88: (The Eulerian) -> brackets needed?
Table 1: ‚Hybrid coefficients‘ are not mentioned nor explained anywhere in the manuscript.
l. 190: Kerkweg, 2006.) -> Kerkweg, 2006).
l. 194: It should be said that C5H8 is isoprene at first occasion.
Fig. 1: In many references to Fig. 1 the word ‚process‘ / ‚processes‘ is missing before the process numbers. E.g., l. 204, 216-217, l. 269, l. 329, l. 356, l. 475 …
l. 266: Maybe rename to multi-phase chemistry to aqueous phase chemistry? Typically multiphase chemistry includes gas phase chemistry, which is not part of MCTS. Further, the ice phase does to have chemical processes except for phase transfer.
l. 306: For this trace species -> For these trace species
l. 334: brackets around ‚the representation of‘
l. 334: microphysic -> microphysical
l. 339: clear -> cloud-free, clear-sky, ..
l. 376-377, l. 489-490: The term 'air mass storm' sounds unfamiliar.
Caption of Figure 3:
- Temperatur -> temperature
- choose -> chosen
- the updraft are for the cloud -> the updraft mass flux are for the cloud
- it’s -> its
Caption of Figure 4:
- Temperatur -> temperature
- choose -> chosen
l. 406: sizes.MCTS -> missing space
l. 406-407: ‚the smaller horizontal domain, by considering horizontal advection instead of subgrid-scale subsidence.‘ -> repetition to l. 404
l. 416: derivations -> deviations
l. 452: The O3 -> The O3 concentration
l. 493: MCTS.This -> missing space
l. 503: similar -> similar to
l. 525: a significant -> significant
Figure 7: I think, plotting the profiles of the reference simulation (Fig. 5) would be helpful for the reader.
l. 570: takes -> takes into account?
Figure 8: Also here, it would be helpful to have some info from the reference simulation to guide the eye, but I understand that it would be somewhat more difficult than for Fig. 7.
Caption of Figure 9:
- tool -> took
- maybe mention date and time of the profile in the model description / setup description
l. 608: ~ above S in ‚Fig. S1‘
l. 636 decease -> decrease
l. 707: is -> are
l. 725: investigates -> investigate
l. 737: composition -> atmospheric composition
l. 805: In the meanwhile, the publication of Barth et al. (2026) is not a preprint anymore. Congrats to your co-author!
Citation: https://doi.org/10.5194/egusphere-2026-4821-RC1
Data sets
Input and Output data MCTS (Mainz Convective Transport and Scavenging (MCTS version 1.0) – Introduction and Evaluation of the new Column Model for Convection-Chemistry-Interactions) Adrienne Jeske, Mary C. Barth, Gustavo C. Cuchiara https://doi.org/10.5281/zenodo.21869008
Model code and software
Mainz Convective Transport and Scavenging model code Adrienne Jeske, Holger Tost, Hendrik Ranocha https://doi.org/10.5281/zenodo.21872890
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 266 | 108 | 45 | 419 | 56 | 57 | 51 |
- HTML: 266
- PDF: 108
- XML: 45
- Total: 419
- Supplement: 56
- BibTeX: 57
- EndNote: 51
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Please find the MCTS code documentation attached. We want to include the MCTS documentation in the supplements in later versions/for following uploads.